druss

I have been looking at trying my hand at breeding some cold hardier bananas. I have collected as many cold hardy wild species as i could. I haven't managed to get musa griersonii or musa darjeeling giant yet and i ruled out musa basjoo as most posts here suggest compatibility issues. I have some musa ornata coming and I have germinated my seeds of Musa sikkemensis, Musa thompsonii, Musa yunnanensis , Musa helens hybrid and Musa red tiger I can look at my plan to do some hybridizing. I am dubious about helens hybrid as no one seems to be able to confirm its parentage.
My initial intent was to cross wild species with diploids. I have Aa: Lacatan, sucrier and inarnibal. Im hoping for ney poovan but as its Ab i feel its usefulness is limited.
As these should carry some or all of the genes needed for parthenocarpy on at least one side of the DNA theres a 50 50 chance of those genes being in any off spring. I will try the pollen cross either way as the parthenocarpy genes could be either side. Im encouraged by the fact there are PNG edible diploid bananas of the phenotype musa AS which are schizocarpa diploids. However if helens hybrid really is a chini champa sikkimensis hybrid then its not guaranteed. i have searched the forum for peoples experiances and
I did read with interest this post that suggested Abb triploids produced fertile pollen, which should carry the parthenicarpic genes:
Weird hybrid
I have Dwarf Ducasse (apparently pisang awak / dwarf namwah?? is this true?), pisang ceylan (mysore), dwarf kalapua, silver bluggoe and blue java. if these produce viable pollen Id like to try pollinating wild diploids to see if I could get triploid seedlings however in this post Gabe stated triploid bananas dont produce viable pollen.
Is the pollen of Saba/cardaba is good for....?
Im going to assume gabe is correct considering his knowledge but are any triploids female fertile? I also have Aaa: red dacca, williams cavendish and gros michele as well as Aab: Pacific plantain, mangaroa torotea and Dwarf french plantain. If the Aab or aaa are female fertile theres a good chance seedling offspring could have the parthenocarpic genes.
More exotic, i have gold finger an Aaab tetraploid, is it female fertile? I could try pollinate it with wild pollen giving possible XAa or xAb genes? I have also been looking into a as home Tc kit for multiplying a few of my edible bananas and I have considered using colchine on wild species such as sikkimensis to create synthetic triploid/tetraploids that could be crossed with edible diploids to perhaps create edible triploids however i have no way to confirm the success or end result of the ploidy effect or if its a stable mutation.
If anyone has actually tried these or has any other suggestions id be open to hear them.

Gabe15

So my post stating that triploids do not have viable pollen is wrong, that was almost 10 years ago and I've learned a lot since then! While it's true that many do not have viable pollen, some do, and a given cultivars ploidy level shouldn't be used to determine it's inherent fertility, either on the female or male side.

Now, that being said, for a simple breeding project I would recommend to avoid using anything with a B-genome at all, as it complicates things a great matter because you have to deal with endogenous BSV (Banana streak virus) inheritance, and there is a tendency for B-containing hybrids to express this later on. There are ways that have been devised to work around this issue, but they are far more complex than what is feasible for a hobby breeding project.

As for transferring parthenocarpy, it is not so simple as a 50/50 chance. It is not a single-gene controlled trait, there at least 3 genes known, and so the chance of any one gene being passed on might be 50%, but the chance of all 3 being passed on to the same seedling is far less (with simple math ignoring other factors it would be closer to 12%), as they are independently sorted. When breeding parthenocarpic to non-parthenocarpic plants, the resulting progeny can have a range of incomplete parthenocarpy if present at all due to only containing some of the genes. The dynamics of how this affects different combinations of varieties is complex, as it greatly depends on which plants you are using, and when using exotic wild species, I would hazard to guess it's an uphill battle as it is less likely those genes are hiding out already in the wild plants, as they are for some wild M. acuminata specimens.

The ploidy levels of resulting progeny from wild x edible crosses can be all over the board, and ultimately are not very important, whats more important is parthenocarpy in a vigorous specimen. There are often hybrid seedlings produced that grow weakly for a time and ultimately end up dying for no apparent reason, and often these are of higher ploidy level than is viable. 2x, 3x and 4x are all capable of being produced and of being viable. 5x, 6x and 7x seedling can also result but generally do not thrive.

However, for all practical purposes, just focus on making crosses with whatever is flowering and see if you get seed, and don't worry too much about the complexities. If you get seed, plant it and see that happens. When working with a small number of plants, the biggest constraint is just having anything flowering at the right time to perform any cross attempt at all, so start there.

One piece of technical advice to start would be to focus on pollinating the wild species with edible pollen if you can. In this way, you will maximize the amount of potential seed you can get as they are all highly female fertile, and if there is any male fertility on the edible variety it would be put to best use on a highly fertile female (i.e., a wild plant). In the reverse sense, using wild pollen on an edible female parent, you may get seed, but most are not very fertile if at all, and the ones that are fertile still are not as fertile as a wild specimen.

Another would be to do everything in the morning as early as you can, at or soon after sunrise. This is when the plants are most fertile and you will have a higher chance of getting seed. Collect male flowers, and immediately pollinate the female flowers. Do not wait until later in the day, or collect flowers the day before for use the next day unless that is the only way you have a chance at making at cross.

__________________
Growing bananas in Colorado, Washington, Hawaii since 2004. Commercial banana farmer, 200+ varieties.

druss

Okay, so reading the above, ideally I should concentrate on pollinating my wild musa species with Aa or Aaa edibles. This should remove the issue of bsv from my offspring and I can focus on growing bigger numbers of those types to increase my chances of synchronous f lowering . Given that I cant find any records of different ploidy levels in wild bananas Im assuming it occurs more commonly in domesicated hybrids. By pollinating wild species with pollen from these groups, using yunnanensis as an example I could theoretically get AY or AaY from the diploid pollen and AY, Aay or AaaY from the triploids? I wouldnt think Id get any AYy or higher levels with the Yy. Once I had these id have to grow them out in isolation until they flowered then:
Check the female flowers werent bisexual, leave 1 bunch to fully flower to see if it was self fertile then grow it amongst others to check if it is partly fertile or sterile.
and finally on the last remove the male flowers to see if it has the p1, p2, p3 or any combo thereof and is hence parthenocarpic. If the fruit fully fill and develop I have a parthenocarpic hybrid of some fertility?

druss

Oh and possibly go collect from the solomon islands, they have lots of Aa types by the looks of it ��

Gabe15

All wild bananas are diploid. There are different numbers of chromosomes between some species, but they are always diploid.

In theory that's true, but it would the same as if you were using M. yunnanensis pollen on a fertile AA or AAA female, as it is a normal wild species with good fertility, it is unlikely to be producing anything other than normal haploid (n) gametes. The 2n restitution gametes that can give double-doses of a genome are not normally occurring in wild species, they are the result of meiotic errors in domesticated clones. The most straightforward way to produce a 2n M. yunnanensis gamete would be create a synthetic tetraploid and it would give 2n gametes.

In reality, it's not really possible to say what you're going to get until you try, and then have some way to analyze the progeny, which is not a simple task. There is also the factor of recombination between genomes which can vary depending on the pairing affinity between the two species which so far is totally unknown, because there has (to my knowledge) been no breeding attempts with M. yunnanensis so far. So if you were to successfully incorporate a Y genome into an AA cultivar for example, it's likely it won't be so clean as being AY or AAY, but rather A(A/Y), or AA(A/Y).

You wouldn't really need to grow the progeny in isolation, you could just bag the female bud to prevent pollen coming from somewhere else. I would think it highly unlikely that a hybrid between M. yunnanensis and an edible variety would be self-pollinating (but if it were that would actually be potentially interesting to generate a new population from), but just in case, you could lift the bracts a day or two in advance of their natural opening and neuter the flowers by clipping the styles and thus preventing any potential accidental pollination. You would then see if the fruits are parthenocarpic or not and to what degree.

__________________
Growing bananas in Colorado, Washington, Hawaii since 2004. Commercial banana farmer, 200+ varieties.

druss

The fact that the wild sp are all diploid should limit the ploidy variation to just the domesticated side. Im hoping to purchase a tissue culture kit (kitchen kulture maybe) to eventually do some cloning . I did consider using this and colchicine to up the ploidy of the wild sp but I have no way of verifying success afterwards. The possibility is there for example to create YyYy tetraploids that could be pollinated to give AYy, AaYy. It would also be useful to incorporate more than one species. I.e. if I crossed griersonii with yunnanensis to get a GY hybrid diploid. If I doubled that to get a GYGY and crossed the result with domestic Acuminatas I could end up with AGY and AaGY. These tetraploids could have benefits as more copies of the wild genes should mean better expression of their traits but it starts to move out of the realm of what I can test and achieve. I think I can realize my goals without going this far. But out of curiosity do labs to the testing on contract? In this vein Im also sticking to the same gene count and section. There are ABT crosses but they dont show any advantages that Im looking for. Also I did consider the rhodochlamys section for shorter cycling and they seem tough but I already have three species, four If I get griersonii and once I start growing enough to try get flowering at the same time thats alot of banana plants.

druss

Okay, sorry to bore everyone �� but I have been considering different approaches to breeding and have come across another bonus (i think) to using only pure A type domestic bananas in my crosses apart from bsv. The nomenclature system used for bananas by Simmonds and sheppard:
Nomenclature of cultivated bananas | News, knowledge and information on bananas
Is based on a scoring system applying points depending on whether certain traits are representative more of acuminata or balbinisia. If I use ab aab abb cultivars in my program, this is of little use. But if I use pure Acuminatas I might be able to copy their system as long as:
A: my wild species (yunnanensis/sikkemensis etc ) show uniformity at the measured points and match taxanomic descriptions.
B: the points of difference used are different between my wild sp and acuminata.
In not sure if anyone has bothered to do this before, but once my wild sp seedlings start to grow I might start taking pics as reference. It will help me confirm my seedlings are what I was told and give me a starting point. The fact I know my wild sp are all diploid as well as the ploidy and consistent A genetics of my cultivated stock should give me a reasonable chance of knowing what my first gen offspring is. I think I 'm going to breed for partheanocarpy first then back breed more to the wild sp. Given the meiosis factor, the A and Y genes will be a bit mixed with the 2nd and subsequent generations. The Y wlll be a bit A and the A will be a bit Y. This doesnt seem to be accounted for in the scoring system? Although it would simply make certain traits appear more one way than the other I guess. Also a call out. Does anyone have a source for griersonii seeds? Im trying kenibreed but its not looking good. Any help is appreciated.